Paper —An Experimental Approach and a Signal Processing Method with the Common Rail Injection… An Experimental Approach and a Signal Processing Method with the Common Rail Injection System of a Diesel Engine

thinquynh@utc.edu.vn Abstract— This paper presents a method and results, which studies ­influences­of­the­fuel­flow­mode­on­the­pressure­oscillation­in­the­volumes­of­ the accumulator fuel system. The fuel is supplied through nozzle holes into a constant­volume­chamber,­which­is­installed­a­drain­orifice­for­fuel­discharge­ into the low- pressure line. Results show that the increase in the base pressure value of the fuel accumulator leads to the rise in the slope of the leading edge of the differential characteristics and the maximum dQ/dt value changes closer to the beginning moment of the fuel injection process. At the same time, the control­pressure­value­is­a­significant­parameter­that­greatly­influences­the­shape­of­the­injection­characteristic.­In­addition,­when­using­the­drain­orifices­with­ different diameters, received values and differential characteristics vary during the fuel supply process. The differential characteristics of the study are the basis for implementing fuel injection control


Introduction
Thedieselengineisstilltrustedbecauseofitshighefficiency,goodeconomy,and large torque, so they are widely used in marine, construction, and agricultural machines, etc. [1][2][3]. In order to optimize the operation process for the highest efficiency, the least fuel consumption, and emission, diesel engines are getting more and more control technologies. The diesel engine control system includes intake air, fuel injection, and emission control, etc. To control those system parameters, the sensors receive the working status of the engine and send it to the ECU for processing. The ECU sends the appropriate control signal to actuators, which depends on the signal received from the sensor.
In addition, studies have demonstrated that injection pressure has a direct effect on the atomizationandmixingefficiencyofdieselengines [4][5][6][7][8].Thefasterspraypenetration contributes to gas utilization and increases combustion rate under high load conditions, and thereby increases engine power [9]. Along with increasing fuel injection pressure, reducing the injector hole diameter helps to reduce soot emissions [10].
In the diesel fuel system, the control process not only is the pressure, the amount of fuel per cycle, but also controls the multi-stage injection as well as with the leading edge of the injection characteristics [11,12]. Moreover, the control process is also involved in the organization and distribution of fuel in the combustion chamber of the engine [13,14]. However, the fuel injection characteristic must be determined under real conditions that develop an appropriate control strategy.
For the above purpose, this paper presents an experimental method to determine the injection characteristics of the common rail fuel system for diesel engines and a method to analyze the signal received from the experiment. The obtained results are the basis for developing a fuel system control system with ultra-high injection pressure (up to 250 MPa and more).

Experimental setup
The test is based on research, that is part of the calculation and experimental complex ICTS-MADI. The installation diagram is shown in Figure 1.
The experimental equipment includes a bed, a CR system, and a microprocessor for the controlling system, which is non-engine tests. The non-engine unit ( Figure 1) has a design modular that allows it to be adapted to the fuel injection process.
To drive the high-pressure fuel pump 3, an electric motor 1 with a converter system 15 is used, which allows speed control. Pressure sensors 6 are installed in the fuel accumulator 5 of the CR system, at the inlet port of the fuel injector, and the chamber 9 for recording the injection characteristics.
During the experiment, the fuel temperature is maintained in the range of (30÷40) o C. To maintain a constant temperature in this range, a fuel cooling system is used. The fuel temperatureintank18isdeterminedthroughatemperaturesensor19. System: 1 -Electric motor; 2 -Shaft coupling; 3 -High pressure fuel pump; 4 -Test bed of the CR system; 5 -Fuel accumulator (rail); 6 -Rail pressure sensors; 7-Electrohydraulicinjector;8-Converterofinjectorcontrolsignals;9-Chamberfor recording injection characteristics; 10 -Low pressure fuel line; 11 -Control pressure gauge; 12 -Liquid measuring cups; 13 -Computer monitor; 14 -Computer cases; 15 -Converter for control the electric motor; 16 -Data acquisition system; 17 -Fuel filter;18-Fueltank;19-Temperaturesensor. Table 1 shows the measurement errors of the pressures recorded during the tests. The fuel system includes a high-pressure pump 3, a fuel accumulator 5, and an electrohydraulic injector (EHI) 7. The nozzle has seven holes with 0.12 mm diameter each. The EHI section is shown in Figure 2. Asafuel,aspecialcalibrationfluid,whichisusedfortestingthedieselfuelsystem. PropertiesofthefluidispresentedonTable2. During the research, signals from pressure sensors, which are installed in the fuel accumulator (p rail ), at the EHI inlet port (p inl ), and in the chamber (p c ) are recorded. At the same time, the signal from the control system to the injector solenoid valve is also written.
The experiment is carried out with a change in the duration pulse (t imp ) from 0.3 to 1.3 ms at three values of the p rail_set pressure in the fuel accumulator, which is set with values of 50, 150, and 250 MPa.
The combined effect study of the p rail_set and the t imp values is performed under the condition of a single injection. Figure 3 shows an example of the current control signal (U set ) and a signal from sensors at the EHI inlet port (U inl ) and signal in the chamber (U c ). The control signal consists of the forming (U set1 ), which ensures the start of the needle, and the holding parts (U set2 ). Throughout all cases, the forming part is 0.3 ms. Ana-lyzingFigure3,sevenstages(from1to7)ofthefuelinjectionprocessareidentified: • Thefirststage.Thisisthestartingofthenozzleneedlemovement,whichischaracterizedbyasignificanteffectofthegapformedbetweentheneedleandthebody nozzle.Theareaoftheeffectiveflowislessthanthatofthesprayholes; • The second stage is the move of the needle. This stage is characterized by an almost linearincreaseinfuelconsumptionwithasignificantdropinpressurep inl ; • The third stage. The p inl value reduces with a stabilization value and it achieves the minimum value due to the movement of the nozzle needle to the maximum position (y = y max ); • The fourth stage. In this stage, the p inl begins of an increase due to replenishment of fuel pressure from the fuel accumulator when the needle is at the highest position y = y max ; • Fifth stage. The nozzle needle falls on the seat, the fuel injection through the spray holes reduces, as a result, the p inl value continues to rise; • Sixth stage. The increase in speed of the p inl value slows down compared to stage 5.Duringthisstage,thecross-sectionofthefuelflowthroughthegapbetweenthe needle and the nozzle body decreases with time. • Seventh stage. At this stage, the p inl pressure value still increases rapidly and the pressurevalueappearswithtwocharacteristicpeaks.Thefirstpeakisaconsequence ofawaterhammerduetoadecelerationofthefuelflowmovingthroughthechannels in the EHI, and the second is a shock when the needle closes the cone of the nozzle.
The drop in the p inl pressure of the first stage takes before the fuel injection start becausethefuelflowfromtheEHIcontrolchamberhasmovedwhentheneedlevalve moves. The time of stage 1 corresponds to the U set1 part of the control signal ( Figure 3). Then, the nozzle needle goes up (starting the U c signal), which is characterized by a certain delay in the U c signal.
As seen from Figure 3, in stages 2 and 3, the change in the p inl value is affected by thefueloutflowthroughthegapbetweentheneedleandthebodynozzlewhentheU set2 pulse of the holding part appears.
The U c signal from the sensor in the chamber illustrates the change of the p c pressure value, which positivelyaffects by the fuel flow from the EHI nozzle holes and negativelyinfluencesbythefueloutflowfromthechamberthroughthedrainorifice.In contrast,thefuelratethroughthedrainorificeisinfluencedbytheeffectiveflowsection and the d j diameterofthedrainorifice.Inaddition,thisflowratealsodependson the pressure drop between the p c values and pressure in the low-pressure line which is determined by the control valve at the end of the line (measured by the control pressure gauge 11 in Figure 1). Figure 4 shows a comparison of the measurement results at two values of the p rail_set pressure (50 and 150 MPa) with different d j values.
With an increase in the value of d j diameter,thefueloutflowratefromthechamber increases, and the p c value decreases during the injection process (comparison between Figure 4, a and Figure 4 For the same modes of the EHI operation (similar t imp and p rail_set ),theflowchange is not observed and its value shall not exceed the error value in the measurement. The p inl pressure value varies with time (p inl = f(t)). In this case, the pc value changes from 5 to 17 MPa in the main injection stage, which depends on selected d j diameter value ofthedrainorifice.Thepcvaluescharacterizethefuelsupplyattheendofthecompression process and operating modes of the diesel engine. Thus, the pressure value in the chamber (p c )hasasmalleffectonthefuelflowfromthenozzleholes,butitisan extremelysignificantroleinthesubsequentdevelopmentofthefuelsprayinthecobustion chamber volume.  Figure 4, d shows the situation of an excessively large diameter of the drain orifice.Atthebeginningofinjection,thefuelrateintothechamberthroughthenozzle holesexceedsitsoutflowthroughthedrainorifice,andthep c pressure in the chamber increases quickly. Further, the p inl pressure decreases (stages 2 and 3 in Figure 3) because the fuel supply rate from the injector decreases and the pressure value between the chamber and the low-pressure fuel line drops, as a result, the p c pressure reduces (Figure 4, d). With an increase in the p inl value in stages 4 and 5 (Figure 3), the fuel rate intothechamberincreases,buttheoutflowfuelratethroughthedrainorificefallsand as a consequence, an increase in the p c valueisseenwith1.2mmofthedrainorifice (Figure 4, d). However, in stage 5 (Figure 3), the nozzle needle moves to the locking cone face of the nozzle body. Thus, the supply fuel amount to the spray holes reduces, and the p c value growth slows down (Figure 4, d).
The throttling of fuel in the gap between the needle and the cone face of the nozzle body in stage 6 ( Figure 3) reduces, which decreases the amount of the injection fuel and as a result, the p c value decreases. After that, the injection process stops, the p c pressure value plunges and is almost linear with time (p c = f(t)), which associates with the fuel flowthroughthedrainorifice.Theslopeofthep c = f(t) line during this period is determined by its hydraulic characteristics.
The variation of the p c = f(t) line at p rail_set = 150 MPa and the different diameters ofthedrainorificeiscomparedbetweenFigure4,dand4,b.Becausethefuelrate is small, the deflection of the p c value is not observed, but there is a decrease with t = (0.7 ÷ 0.9) ms (Figure 4, b).
Comparing Figure 4, a and Figure 4, b, as well as Figure 4, c and Figure 4, d, it can see that the decrease in p c = f(t) value starts at the same point after the injection process stops at the same operating modes of the EHI (p rail_set , t imp ) with different d j value.
Basedontheanalysisoftheobtainedresults,amethodisformulatedfordetermining the differential injection characteristic. According to this method, every moment t, the pressure rate in the combustion chamber (dp c /dt) and the pressure rate through the drain orifice(dp j /dt) is shown in formula 1: The (dp j /dt) value is determined for the p c value along with the failing line p c = f(t) aftertheinjectionprocessstopswithonlyfuelflowthroughthedrainorificewhichrepresentsthehydrauliccharacteristicofthedrainorifice.Figure4showsthatfordifferent t imp values and certain values of the p rail_set and the d j , the descending line parts of the p c = f(t) evenly space. In this regard, it is necessary to perform an additional registration of the p c = f(t) at the t imp that exceeds the maximum value of control duration pulse when the injection characteristics in the change range of the t imp value are performed.
When the method is implemented for determining the differential injection characteristics through the formula S(dр/dt) = f(t), it is necessary to select the d j diameter value, which ensures a monotonous increase in the p c pressure (except for the end of stage 6, Figure 3) during the injection process, as shown in Figure 4, a and Figure 4, c.
The conversion S(dр/dt) = f(t) in the differential characteristics of the injection dQ/dt=f(t)ismadethroughacoefficientwhichrepresentsthrougharatiobetweenthe area limited S(dр/dt) = f(t) and the amount of injection cycle (Q c ) in the fuel injection cases (p rail_set , t imp ) ( Figure 5).  Figure 6 shows p inl = f(t) in the fuel injection process, which depends on the duration t imp of the control pulse. In this case, Figure 7 illustrates the comparison between the pressure change at the injector inlet port (Dр inl ) with different p rail_set values and three t imp values 0.3, 0.7, and 1.3 ms.
With an increase in the p rail_set the fuel flow rate through the spray holes goes up, which leads to a drop in the p inl value with a greater Dp inl value at stage 2 of the injection process when the needle is lifted (Figure 7). At the same time, the higher values of the fuelflowrateandthep inl valueare(Figure6),thebiggerthefuelflowrategets.This leads to an increase in the Q c value at the same t imp value.
When the fuel supply rises with a growth in the p rail_set , the Dp inl value drop in the next stages (3÷6) of the injection processes also goes up (Figure 7). Anincreaseinthefuelflowratewithariseinthep rail_set leads to the release of more energy when it is decelerated (water hammer). The result is a growth in the p inl value of thefirstpeakinstage7(Figure7).Inthiscase,becausetheabsolutevaluep inl at the end ofthecontroldurationpulseisbigandtheEHIvalveisclosed,thereforethefuelfilling rate in the control chamber of the nozzle is fast, as a consequence, the fall speed of the needle on the seat increases. The energy in the fall process of the needle rises, which is shown an increase in the p inl value of the second peak at stage 7 ( Figure 7). This is especially noticed with a rise in the t imp (Figure 7, c).
Besides the intensity of the fuel flow through the nozzle holes, the p inl value is affectedbythefueloutflowfromtheEHIcontrolchamberwhenthevalveopens.The early termination of the control pulse (the decrease in the t imp ) contributes to a decrease in the p inl droprate,butthiseffectishardlynoticeablebecausetheeffectiveflowsectionofthenozzleislargerthantheflowthroughthegapbetweentheneedleandnozzle body. However, the lower the t imp value and the higher the p inl pressure are, the higher thefillingrateinthecontrolchambergetsandalsothehighertheneedlespeedfallson the seat. This is accompanied by a growth in the value of the second p inl peak in stage 7 ( Figure 6, b and Figure 6, c). The high needle falling speed on the seat causes a rise in flowdecelerationandanincreaseinthep inl valueofthefirstpeakinstage7(Figure6,b and Figure 6, c). In this case, an increase in the t imp value contributes to a rise in the deflectionofthep inl pressure value between the two peaks in stage 7 ( Figure 6).
The relationship between the decrease in the t imp value and the increase in the falling speedoftheneedleontheseatisinfluencedbythep rail_set value. Growth in the p rail_set value contributes to a drop in the p inl in stage 2 of the injection process. This is clearly seen from the comparison of Figure 6, a (p rail_set = 50 MPa) with Figure 6, b (p rail_set = 150 MPa) and Figure 6, c (p rail_set = 250 MPa). At p rail_set = 50 MPa and t imp = 0.3 ms, the p inl value of the pressure peaks in stage 7 are even lower than it at t imp = 0.5 ms.
The results of the received signals processing from the sensors with the formula p inl =f(t)andmassflowratedQ/dt = f(t) and the control pulse at the p rail_set values equal to50,150,and250MPaareshowninFigure8,Figure9andFigure10,respectively.
The increase in the p rail_set value and its effect on the p inl change is described above, which leads to a rise in the slope of the leading edge of the differential injection characteristic and the maximum dQ/dt changes closer to the beginning of the fuel injection (Figure8,Figure9andFigure10).Thisproblemisexplainedbythedecreaseinthe intensity of fuel supply which is represented by the fall in the p inl value. The middle and at the end of the fuel injection process are affected by the change in the p inl value in stages (4 ÷ 6). With an increase in the p rail_set value, more fuel is injected at the moment of falling rate(comparisonofFigure8withFigure9andFigure10).Inthiscase,thehigherthe p rail_set amount is, the bigger the intensity value of the dQ/dt drops from the maximum value to the end of injection (comparison of Figure 9 and 10). Thus, the p rail_set control valueisasignificantparameterwhichinfluencestheshapeoftheinjectioncharacteristic. It also redistributes the fuel amount which is supplied by the fuel injection process by changing the injection duration.
It is obvious that the studied effects on the p inl change ( Figure 6 and Figure 7) have a combined effect on the p rail_set , t imp , the speed of the control valve, the volume of the controlchamber,theeffectiveflowsectionofthenozzle,andtheinternalvolumeinthe EHI body and the move of the nozzle needle to the cone face.

Conclusion
Qualitative and quantitative analysis of the research results suggests the conclusions that a method is developed to determine the differential characteristics of the fuel injection by summing at each moment t with the change rate of pressure in the chamber (dр c /dt)andthefallrateofpressureduetothefueloutflowthroughthedrainorifice (dp j /dt). The increase in the base pressure value of the p rail_set fuel accumulator leads to the rise in the slope of the leading edge of the differential characteristics and the maximum dQ/dt value changes closer to the beginning moment of the fuel injection process. The p rail_set controlvalueisasignificantparameterthatinfluencestheshapeof the injection characteristic. It also redistributes the fuel amount which is supplied by thefuelinjectionprocessbetweenthestartingandthefinalstageswithchangingofthe injection duration.